PROJECT SUMMARY/ABSTRACT Advanced heart failure is characterized by progressive debilitating symptoms and repeated hospitalizations that degrade quality of life. There is no one criterion to diagnose advanced heart failure; the definition is complex and challenging to apply broadly to populations. As such, our knowledge of advanced heart failure is truncated and skewed as it is based on information from referral populations and convenience samples. Enhancing our understanding of the epidemiology, experiences, and outcomes of patients with advanced heart failure is critical to developing interventions to improve care and quality of life. To address these gaps in knowledge, this proposal leverages diverse data sources and novel applications of quantitative and qualitative methods to assess the epidemiology and outcomes of individuals with advanced heart failure. In Aim 1, we will apply an advanced heart failure definition to a geographically-defined population of individuals with heart failure under the auspices of the Rochester Epidemiology Project. We will determine the prevalence of advanced heart failure, examine the demographic and clinical features of the population, and evaluate the timing of its development and association with risk of outcomes. In Aim 2, we will use machine learning techniques to develop computer algorithms (computable phenotypes) to identify patients with advanced heart failure using electronic health record data. We will then leverage the infrastructure of the National Patient- Centered Clinical Research Network (PCORnet) to validate the performance of the computable phenotypes across diverse patient populations. This will enable the accurate and efficient identification of advanced HF for future applications. In Aim 3, we will use the computable phenotype developed in Aim 2 to prospectively identify individuals living with advanced HF. We will then assess their treatment and illness burdens using a combination of surveys and semi-structured qualitative interviews. This information will be used to inform the development of a palliative care intervention that is tailored to the needs of patients with advanced HF. We will assess the acceptability of the tailored palliative care intervention to stakeholders (patients, caregivers, clinicians). The results of these analyses will provide synergistic information to clarify the epidemiology, case mix, burdens, and outcomes of individuals with advanced heart failure. They will provide a prototype palliative care intervention tailored to decrease burden and improve quality of life in advanced heart failure. Finally, the computable phenotype developed can be used to identify patients with advanced HF for future quality improvement programs, observational studies, and interventional research.